Jean-François Baffier
Research Subjects
Jean-François Baffier has a total of 3 co-authors. Here are the top co-authors by number of shared publications:
- Florian Richoux
AIST
2 shared publications - Florian Richoux
AIST Japan
1 shared publication - Philippe Codognet
JFLI, CNRS/Sorbonne University/University of Tokyo
1 shared publication
Latest Publications
- 26 June 2023
- Published by Springer Nature
- p. 153-167
Learning qubo Models for Quantum Annealing: A Constraint-Based Approach
- 26 June 2023
- Published by Springer Nature
- p. 153-167
Learning qubo Models for Quantum Annealing: A Constraint-Based Approach
- Open Access
- 20 February 2023
- Published by Springer Nature in Annals of Mathematics and Artificial Intelligence
- Vol. 93 (3) , 441-469
Automatic error function learning with interpretable compositional networks
- 20 February 2023
- Published by Springer Nature in Annals of Mathematics and Artificial Intelligence
- Vol. 93 (3) , 441-469
Open AccessAutomatic error function learning with interpretable compositional networks
- Open Access
Learning Interpretable Error Functions for Combinatorial Optimization Problem Modeling
In Constraint Programming, constraints are usually represented as predicates allowing or forbidding combinations of values. However, some algorithms exploit a finer representation: error functions. Their usage comes with a price though: it makes problem modeling significantly harder. Here, we propose a method to automatically learn an error function corresponding to a constraint, given a function deciding if assignments are valid or not. This is, to the best of our knowledge, the first attempt to automatically learn error functions for hard constraints. Our method uses a variant of neural networks we named Interpretable Compositional Networks, allowing us to get interpretable results, unlike regular artificial neural networks. Experiments on 5 different constraints show that our system can learn functions that scale to high dimensions, and can learn fairly good functions over incomplete spaces.
Open AccessLearning Interpretable Error Functions for Combinatorial Optimization Problem Modeling
In Constraint Programming, constraints are usually represented as predicates allowing or forbidding combinations of values. However, some algorithms exploit a finer representation: error functions. Their usage comes with a price though: it makes problem modeling significantly harder. Here, we propose a method to automatically learn an error function corresponding to a constraint, given a function deciding if assignments are valid or not. This is, to the best of our knowledge, the first attempt to automatically learn error functions for hard constraints. Our method uses a variant of neural networks we named Interpretable Compositional Networks, allowing us to get interpretable results, unlike regular artificial neural networks. Experiments on 5 different constraints show that our system can learn functions that scale to high dimensions, and can learn fairly good functions over incomplete spaces.